Automatic integration with LLM tools

## Documenting usage of llms.md

As a maintainer of developer tooling (https://github.com/a-h/templ) I've had an issue raised to implement this.

However, documenting how developers should use `llms.md` is a burden on me. I use https://github.com/CopilotC-Nvim/CopilotChat.nvim so I can document how to load `llms.md` into the context, but I have no idea how to configure other LLM coding tools.

More and more LLM tools will be created over time, so it's a bit of a losing battle. If this project documented how to use it, it would save maintainers time.

## Automatic use of llms.md by tooling

If a Go developer is using `github.com/a-h/templ` in their project, then their LLM coding tool should be able to automatically pick up that it should probably load my context.

So, actually, this standard should probably include having a `.llm-context.json` file or similar in git projects that maps filetypes to context locations that all LLM tools can use, e.g.:

```json
{
"*.templ": [ "https://templ.guide/llms.md" ]
}
```

Of course, it would be a security risk, since you'd be forcing users to download files, and be able to collect their IP addresses and other information, so the LLM tooling UX would have to ask first before downloading.

No repository files or tests are named; start by reviewing the llms.md usage described here and the proposed .llm-context.json mapping. Define the supported tooling, download-consent behavior, and acceptance criteria for automatic context loading before implementation. Done should include documented usage and an agreed security scope.

Written by the indexing model from the issue text.